Why Pay Taxes? A Review of Tax Compliance Decisions Erich

International Studies Program

Working Paper 07-30

December 2007

Why Pay Taxes? A Review of Tax

Compliance Decisions

Erich Kirchler

Stephan Muehlbacher

Barbara Kastlunger

Ingrid Wahl

International Studies Program

Andrew Young School of Policy Studies

Georgia State University

Atlanta, Georgia 30303

United States of America

Phone: (404) 651-1144

Fax: (404) 651-4449

Email: [email protected]

Internet: http://isp-aysps.gsu.edu

Copyright 2006, the Andrew Young School of Policy Studies, Georgia State University. No part

of the material protected by this copyright notice may be reproduced or utilized in any form or by

any means without prior written permission from the copyright owner.

International Studies Program Working Paper 07-30

Why Pay Taxes? A Review of Tax Compliance Decisions

Erich Kirchler Stephan Muehlbacher Barbara Kastlunger Ingrid Wahl

December 2007

International Studies Program Andrew Young School of Policy Studies The Andrew Young School of Policy Studies was established at Georgia State University with the objective of promoting excellence in the design, implementation, and evaluation of public policy. In addition to two academic departments (economics and public administration), the Andrew Young School houses seven leading research centers and policy programs, including the International Studies Program. The mission of the International Studies Program is to provide academic and professional training, applied research, and technical assistance in support of sound public policy and sustainable economic growth in developing and transitional economies. The International Studies Program at the Andrew Young School of Policy Studies is recognized worldwide for its efforts in support of economic and public policy reforms through technical assistance and training around the world. This reputation has been built serving a diverse client base, including the World Bank, the U.S. Agency for International Development (USAID), the United Nations Development Programme (UNDP), finance ministries, government organizations, legislative bodies and private sector institutions. The success of the International Studies Program reflects the breadth and depth of the in-house technical expertise that the International Studies Program can draw upon. The Andrew Young School's faculty are leading experts in economics and public policy and have authored books, published in major academic and technical journals, and have extensive experience in designing and implementing technical assistance and training programs. Andrew Young School faculty have been active in policy reform in over 40countries around the world. Our technical assistance strategy is not to merely provide technical prescriptions for policy reform, but to engage in a collaborative effort with the host government and donor agency to identify and analyze the issues at hand, arrive at policy solutions and implement reforms. The International Studies Program specializes in four broad policy areas: Fiscal policy, including tax reforms, public expenditure reviews, tax administration reform Fiscal decentralization, including fiscal decentralization reforms, design of intergovernmental

transfer systems, urban government finance Budgeting and fiscal management, including local government budgeting, performance-

based budgeting, capital budgeting, multi-year budgeting Economic analysis and revenue forecasting, including micro-simulation, time series

forecasting, For more information about our technical assistance activities and training programs, please visit our website at http://isp-aysps.gsu.edu or contact us by email at [email protected].

Why pay taxes? A review of tax compliance decisions 1

Why pay taxes? A review of tax compliance decisions

Erich Kirchler*, Stephan Muehlbacher, Barbara Kastlunger, Ingrid Wahl

*Corresponding Author:

University of Vienna, Faculty of Psychology

Universitaetsstrasse 7, 1010 Vienna, Austria

E-mail: [email protected]

T: +43 1 4277 47880, F: +43 1 4277 47889


Abstract

Since the standard model of income tax evasion (Allingham & Sandmo, 1972; Srinivasan,

1973) was published, much research has tested its four parameters (level of actual income,

tax rate, audit probability, penalty rate) for empirical validity. Surveys, laboratory

experiments and analysis of aggregate data revealed ambiguous evidence for the model’s

behavioral implications. The present article reviews these studies and concludes that

compliance decisions can only partly be explained by the rational choice approach. We

suggest that depending on the climate in a society, compliance stems from two different

factors. In a climate of distrust, high power of authorities is needed to enforce tax

compliance and increasing fines and audit probabilities may be an effective tax policy. In a

climate where taxpayers trust the authorities of their state, however, other variables gain in

importance. Knowledge, attitudes, moral appeals, fairness and democracy may lead to

voluntary compliance. In this case, draconian fines and intrusive audits can take

unintentional effects and would corrupt tax morale.

Keywords: compliance decisions, tax evasion, audits, fines, trust


WHY PAY TAXES? A REVIEW OF TAX COMPLIANCE DECISIONS

1. Introduction

Tax compliance has evolved into a major research topic in economic psychology. The

issue has been approached from various viewpoints shedding light on different aspects of

taxpayers’ behavior. Attitudes were measured, prevailing social norms captured and lay

theories explored, which people have in mind when brooding over their annual tax

declarations (for an overview see Kirchler, 2007).

One line of research focuses on judgment and decision processes in tax compliance. In

the seminal works of Allingham and Sandmo (1972) and Srinivasan (1973) income tax

evasion was modeled as decision under uncertainty1. Since then, parameters specified in

the formal model were empirically studied in various publications. This review summarizes

results and conclusions of research on tax compliance decisions within the paradigm of

maximizing expected utilities. After illustrating the model, each of its parameters is

addressed separately and empirical evidence for or against their effects on tax compliance

is reported. Finally, a summary of results is provided, typical problems of tax compliance

research and alternative theoretical approaches are discussed.

Allingham and Sandmo’s (1972) and Srinivasan’s (1973) analyses are restricted to

income tax evasion. Research dealing with value added tax, other taxes or related duties,

therefore, is omitted in the present review. Most of the cited studies focus on self-employed

taxpayers, because this group has the most opportunities to evade income taxes.

2. Tax evasion as a decision under uncertainty

Following the neoclassical approach in economics, Allingham and Sandmo (1972) and

Srinivasan (1973) assume that taxpayers are rational agents whose choice behavior

conforms to the Von Neumann-Morgenstern axioms, and who try to maximize utility of

1 In fact, Allingham and Sandmo (1972) and Srinivasan (1973) published their analyses of tax evasion nearly at the same time without knowing about each other’s work.


their taxable income. They do so by weighing costs and benefits of compliance with the

expected utility of evasion. Choosing the risky strategy of evading all or part of actual

income will result in a better outcome, if the respective tax file is not audited. In case of an

audit, however, compliance would have been the better strategy since paying a fine

decreases income even further than an honest payment of one’s taxes.

Accordingly a compliance decision depends on (i) the level of actual income, (ii) tax

rates, (iii) audit probabilities and (iv) the magnitude of fines. Though Allingham and

Sandmo (1972) admit that other, less economic variables might be also important in

understanding tax compliance (and exemplarily outline a more complex model

incorporating changes in the taxpayer’s reputation among society due to his or her

compliance behavior), their analysis is based on the simpler model with the four parameters

described above.

Since their analysis – often (and herein after) referred to as the standard economic

model - was published, empirical research gathered evidence for (and against) the

significance of its parameters in tax compliance decisions. The following four sections give

a review of these studies, structured by the respective variable being tested. Each section

begins with predictions of the formal model and concludes in potential explanations for

empirical deviations. Implications for tax policy and everyday practice of fiscal authorities

are discussed.

3. Level of actual income

Regarding income-effects, predictions of the standard economic model are ambiguous.

Though wealthier citizens are more likely to evade their taxes since the level of absolute

risk aversion decreases with income, it is not clear if severity of evasion is an increasing or

decreasing function of income. Once taxpayers have decided to evade taxes, the degree of

underreporting depends on their relative risk aversion (i.e., the risk-attitude at a specific


point of the utility function). The relation of income and relative risk aversion is not unique

and consequently Allingham and Sandmo (1972) conclude “…that, when actual income

varies, the fraction [of income] declared increases, stays constant or decreases according as

relative risk aversion is an increasing, constant or decreasing function of income” (p. 329).

As ambiguous as the standard economic model’s predictions is the empirical evidence

for income effects on compliance. If any impact of income level was found, both proposed

directions of the effect were supported.

A negative relation of income and tax compliance is reported by Slemrod (1985), who

analyzed archival data from the United States’ Treasury tax file for 1977, and by Ali, Cecil

and Knoblett (2001), who analyzed IRS data for the period between 1980 and 1995.

Consistently, Weck-Hannemann and Pommerehne (1989) found lower compliance among

high income earners in archival data on Swiss taxpayers. Lang, Nöhrbaß, and Stahl (1997)

combined survey data from 33 000 West German households with national accounting

data. Again, wealthier citizens were shown to be particularly prone to evade and avoid

taxes. Further evidence for a negative relation is reported from laboratory experiments.

Baldry (1987) manipulated participants’ income (among other variables) and found lower

compliance at higher income-levels. Anderhub, Giese, Güth, Hoffmann, and Otto (2001)

report similar results from a multi-period experiment, where participants had to earn their

income in small tasks. Collins and Plumlee (1991) also found decreased compliance among

participants with higher income. The importance of how one’s economic status is

perceived, was demonstrated by Vogel (1974). Taxpayers who reported improvement of

their economic status were less compliant than others who reported deterioration of their

financial well-being.

A positive relation of income and tax compliance also found empirical support.

Christian (1994) analyzed data of the Taxpayer Compliance Measurement Program


(TCMP) of the Internal Revenue Service (IRS) in the United States. On average, taxpayers

with an annual income of more than $ 100 000 reported 96.6% of income, but only 85.9%

if income was below $ 25 000. Fishlow and Friedman (1994) found decreased compliance

at low income-levels in archival-empirical data from Argentina, Brazil and Chile, three

countries with low economic growth and high inflation rates. A single laboratory

experiment found also support for a positive relation of income and compliance. Alm,

Jackson, and McKee (1992) report that in their experiment compliance significantly

increased with income.

Other studies found no relation of income level and tax compliance. Feinstein (1991)

compared pooled data from the 1982 and 1985 TCMP and found no significant effects of

income on tax compliance. In a laboratory experiment conducted in Korea, Park and Hyun

(2003) among other variables varied experimental income. No effect of income on

compliance behavior was found. Self-reported compliance behavior was not related to

income among American (Porcano, 1988) and Swedish taxpayers (Wärneryd & Walerud,

1982).

Definite conclusions can not be drawn from the empirical studies reported. Evidence for

income effects is inconsistent (as Andreoni, Erard, and Feinstein, 1998, already concluded

in an earlier review). Though a slight majority of studies reports a negative relation of

income and compliance, also opposite and zero effects were found.

The ambiguous evidence could possibly be explained by other variables moderating or

mediating the effect of income. Besides the problem of relative risk-aversion, several

variables can be imagined to weaken the income effect or to trigger its direction. For

instance, opportunities for tax avoidance might vary with income. High-income earners are

able to afford professional tax advisors, who know about the loopholes in tax law

(Slemrod, Blumenthal, & Christian, 2001; Wärneryd & Walerud, 1982). If tax avoidance is


possible, the tax burden can be reduced without breaking the law and taking the risk of

paying a fine. Spicer and Lundstedt (1976) point out that also self-employed have more

possibilities to avoid taxes than employed taxpayers. However, self-employed taxpayers

have also more opportunities for tax evasion and opportunities might further increase with

the number of different income sources. Hence, in compliance decisions the level of

income might interact with its source. A different aspect of the income source - if income

was earned by hard work or an effortless job - has been studied in experiments by Kirchler,

Muehlbacher, Hölzl, and Webley (in press; Muehlbacher & Kirchler, in press). Participants

were less compliant when they reported income earned by low effort than when they

reported hard-earned income. It seems that taxpayers are reluctant to lose their hard-earned

money by “gambling” with tax authorities.

Given the present results on income and compliance, it is too early to consider practical

implications for tax policy. The issue, however, is important not only for designing proper

audit strategies, but also for tax ethics, since a negative relation of income and compliance

would put “[…] into question the intended (or pretended?) distributional effects of

progressive income taxation.” (Lang, Nöhrbaß, & Stahl, 1997, p. 328)

2. Tax rate

Would higher tax rates decrease compliance? No clear hypothesis emerges from the

standard economic model. Two counteracting effects are proposed: On the one hand, a high

tax rate reduces effective income and, therefore, makes tax evasion more profitable. On the

other hand, by reducing effective income absolute risk aversion increases. Consequently,

evasion should be reduced.2

Both of the model’s predictions on the impact of tax rates on compliance found

empirical support, though most studies report that compliance is lower at high tax rates. 2 The latter proposition is also supported by Yitzhaki (1974), who suggests that if fines are posed on the evaded tax instead of undeclared income, compliance will be related positive to tax rates.


A negative relation of tax rate and compliance was found by Clotfelter (1983), who

analyzed American taxpayers’ data from the 1969 TCMP, and by Slemrod (1985), who

studied data from the United States’ Treasury tax file of 1977. Consistently, lower

compliance at high marginal tax rates is reported by Lang, Nöhrbaß, and Stahl (1997) for

German taxpayers and by Pommerehne and Weck-Hannemann (1996; Weck-Hannemann

& Pommerehne, 1989) for Swiss taxpayers. An analysis of IRS data for the period between

1980 and 1995 revealed that the negative effect of tax rate on compliance is more

pronounced among high-income earners (Ali, Cecil, & Knoblett, 2001). Experimental

evidence for the negative impact of high tax rates is reported by Alm, Jackson, and McKee

(1992). In 25 periods participants received an income, paid taxes and were randomly

audited and fined. During the sessions of the experiment tax policy changes were

introduced, including a change of tax rate (10%, 30%, and 50%). When tax rate increased

in the experiment, participants’ compliance decreased. Similar results are reported by

Friedland, Maital and Rutenberg (1978) from an experiment where tax rate varied between

subjects (25% and 50%). Thus, instead of the impact of an increasing or decreasing tax

burden, this experiment tested for the effect of absolute differences in tax rates. Actually

studying the effect of different audit schemes, Collins and Plumlee (1991) tested also for

the effect of different tax rates (30% vs. 60%). Again, lower compliance was observed at

the higher tax rate. An experiment by Moser, Evans III, and Kim (1995) revealed the

importance of perceived fairness when studying the impact of different tax rates. A

negative effect of a high tax rate was found only if participants felt inequitably treated

compared to others.

A positive relation of tax rates and compliance, as proposed by Yitzhaki (1974), is less

often reported. Feinstein (1991) found higher compliance at higher tax rates in aggregate

data on American taxpayers from the TCMP. In an experiment Alm, Sanchez, and De Juan


(1995) conducted in Spain, participants were more compliant when tax rate increased over

time.

No effect of the tax rate was found in an experiment by Baldry (1987) and in a study by

Porcano (1988), dealing with self-reported compliance behavior.

To summarize, most empirical studies on the impact of tax rates support the assumption

that high tax burdens have negative impact on compliance. However, the strong connection

of income and tax rate makes final conclusions difficult (Andreoni, Erard, & Feinstein,

1998; Slemrod, 1985). In experimental studies it is hard to separate the effects of tax rates

and income, if both variables are varied at the same time. In field studies a similar critique

as in the discussion on income effects can be made, i.e., opportunities for tax evasion or

avoidance and source of income are likely to interact with tax rate. Regarding the source of

income, Boylan and Sprinkle (2001) report that participants in their experiment reacted to a

tax rate increase with lower compliance if they were endowed with income by the

experimenters, but reacted with higher compliance to a tax rate increase if they had to earn

their income by performing in a one-hour multiplication exercise. Findings by Moser,

Evans III, and Kim (1995) seem to suggest that perceived fairness of the tax rate is more

important than its absolute level. Judging the fairness of taxation, however, requires

comprehensive knowledge and correct interpretation of the tax law, and complex tax rate

structures, such as progressive taxation, are not perfectly understood by taxpayers. Roberts,

Hite, and Bradley (1994) showed that taxpayers’ preferences for progressive, flat, and

regressive taxation depend on the form the respective tax rates are presented, e.g., if tax

rates are described as abstract concept or in concrete terms by giving hypothetical

examples. McCaffery and Baron (2004) asked their participants to indicate what they

consider a fair tax for different levels of income. Preference for progressive taxation was

stronger when the amount of taxes had to be indicated in percentages of income than when


it had to be indicated in absolute amounts. Such framing effects suggest that tax policies

are hard to understand and can easily be misinterpreted. Knowledge about framing effects

could help authorities in promoting changes in tax policy such as tax rate increases without

undermining taxpayers’ compliance.

3. Audit probabilities

The economic model assumes that taxpayers try to maximize the outcome of the

compliance decision by weighing the gain of successful evasion against the risk of

detection and punishment. The expected value of non-compliance depends on audit

probabilities and fines. Allingham and Sandmo (1972) therefore conclude that “… an

increase in the probability of detection will always lead to a larger income being declared”

(p. 330).

Most empirical studies found support for the effect of audit probability on compliance,

though the reported effects are sometimes weak (c.f., the specific review on audit rates by

Fischer, Wartick, & Mark, 1992).

A positive effect was found in aggregate data on tax compliance behavior of Swiss

taxpayers. Pommerehne and Weck-Hannemann (1996; Weck-Hannemann & Pommerehne,

1989) compared 25 Swiss cantons and report that compliance was higher in cantons where

more audits occurred in the observed period. Ali, Cecil, and Knoblett (2001) report similar

results in an analysis of a dataset from the American IRS. However, audit probability

affected compliance more among high-income earners. An opposite interaction of audit

probability and income was found in a field experiment with a sample of 1 724 American

taxpayers by Slemrod, Blumenthal, and Christian (2001). The experimental group in this

study received a notification by mail that their next tax files will be audited. While low and

middle-income earners increased tax payments compared to previous years, high income

earners reacted differently and indicated a lower tax liability. The authors assumed that the


notification letter induced high income taxpayers to seek help of professional tax advisors

to minimize their tax liability. The effectiveness of frequent audits found also support in

laboratory experiments. Alm, Sanchez, and De Juan (1995) compared audit rates of 5%,

30%, and 60% and found a corresponding increase in compliance. In the experiments by

Trivedi, Shehata, and Lynn (2003) and Trivedi, Shehata, and Mestelman (2004) audit rates

were either zero or 25%. As expected, compliance was higher in the latter condition.

Gërxhani and Schram (2006) conducted experiments in Albania and in the Netherlands.

Participants in Albania were not affected by audit rates (16.6% and 50%), but Dutch

participants evaded more when audit probability was low. That the relation between audit

probability and compliance is non-linear revealed an experiment by Alm, McClelland, and

Schulze (1992). At audit probabilities of 0%, 2%, and 10% participants declared 20%,

50%, and 68% of their income. Spicer and Thomas (1982) varied audit rates and provided

their participants with either concrete (1 in 20, 5 in 20, and 3 in 20) or non-concrete

information (low, high, and medium) on the probability of being audited. Again, with

increasing audit rates evasion declined. Furthermore, precise information had stronger

impact than imprecise information. By contrast, Friedland (1982) reports that at small audit

probabilities imprecise information can enhance compliance.

Fischer, Wartick, and Mark (1992) point out that perceived and objective probabilities

widely differ. Given the extreme low audit probabilities in most countries, tax evasion

should approach 100% if no overweighting occurs as suggested in prospect theory

(Kahneman & Tversky, 1979). Subjective rather than objective probabilities should

therefore be of greater interest in tax research. Survey research revealed that taxpayers who

admitted to evade taxes perceive chances of being caught lower than honest taxpayers

(Mason & Calvin, 1978). However, correlations of perceived risk and compliance are

sometimes weak or even not present (Elffers, Weigel, & Hessing, 1987; Spicer &


Lundstedt, 1976; Wärneryd & Walerud, 1982). Subjective probability that an audit will

occur may partly depend on prior experience. In an experiment by Spicer and Hero (1985)

participants paid taxes on a given income and were randomly checked in overall 10

periods. Compliance in the tenth period was correlated with the number of audits in

previous rounds. Similar results from a business simulation study are reported by Webley

(1987). That the experience of an audit induces a learning process for evaluating audit

probabilities is suggested by findings from Guala and Mittone (Guala & Mittone, 2005;

Mittone, 2006). Participants in their experiments paid taxes in overall 60 periods and were

audited either only in the first or only in the last 30 periods. If participants were audited

early in the experiment, compliance was relatively high and remained at a high level also in

the latter 30 periods, though no more audits had occurred. By contrast, if participants were

not audited in the first half of the experiment, compliance was comparably low and did not

increase even when audits started in the latter 30 periods. Learning and understanding

objective probabilities of uncertain events by experiencing (or observing) their occurrence

can be understood as applying the availability heuristic (Tversky & Kahneman, 1974).

Analogous to Tversky and Kahneman’s example that seeing a burning house influences

judgments on the occurrence of such accidents, experiencing an audit could increase the

subjective probability of being audited. On the other hand, it was frequently observed in tax

experiments that compliance sharply decreases after an audit (Guala & Mittone, 2005;

Mittone, 2006). Such a reaction to an audit has been coined “bomb-crater” effect by

Mittone (2006). In a recent study Kastlunger, Kirchler, Mittone and Pitters (submitted

manuscript) investigated if the “bomb-crater” effect is caused by misperception of chance

or by the attempt to repair losses from paying a penalty. Results suggest that misperception

of chance is the major cause for the strong decrease in compliance immediately after an

audit. At least in some cases, however, also loss repair tendencies seem to reduce


compliance. The “bomb-crater” effect affects only the first periods after an audit before

compliance increases again. After a “bomb-crater” occurred in an experiment by Kirchler,

Maciejovsky, and Schwarzenberger (2005) compliance increased faster to its baseline in a

condition with a high audit rate (30%) than in a condition with a lower audit rate (15%).

Thus the duration of this effect seems to depend on when the next audit is expected.

To summarize, empirical evidence for the impact of high audit probabilities is quiet

strong. Yet, in practice tax audits are costly and research consequently changes its focus

towards alternative control mechanisms. Reinganum and Wilde (1985), and Alm,

Cronshaw, and McKee (1993) contrasted random audits schemes with several alternatives.

They tested the effectiveness of a cut-off rule (i.e., when the declaration falls below a

certain threshold), a retrospective audit scheme (i.e., when the random detection of non-

compliance results in examination of previous tax files), and a prospective audit scheme

(i.e., when the detection of non-compliance increases future audit probability). Alternative

audit mechanisms lead to higher tax compliance even if the audit rules implicated fewer

audits than random control systems. Also the findings by Guala and Mittone (Guala &

Mittone, 2005; Mittone, 2006) can be used to design an efficient audit-scheme. According

to their results it might be advisable to audit particularly young and inexperienced

taxpayers. Probably they will “learn” to be compliant if their very first tax file is checked

by authorities.

4. Fines

The second parameter determining the expected value of being non-compliant besides

audit probabilities is the amount of fines. Due to their multiplicative linkage fines and audit

rates may substitute each other, as long as neither of them is set to zero. Higher fines

simply make evading taxes more hazardous for taxpayers and should, therefore, deter from

evasion.


Empirically the deterrent effect of fines could not always be supported. The observed

effects were weaker than expected and some studies even suggest that an increase of

penalties can have undesirable effects and result in more tax avoidance.

Supporting evidence for the effect of fines is reported by Alm, Jackson, and McKee

(1992), though its impact on compliance was virtually zero. In experiments by Friedland,

Maital, and Rutenberg (1978), and by Park and Hyun (2003) compliance was stronger

affected by the amount of fines than by audit probabilities. By contrast, Friedland (1982)

reports that audit rates affected compliance more than the amount of fines. Finally, Alm,

Sanchez, and De Juan (1995) point out that fines are only effective in combination with

high audit rates. The interaction of both variables seems to be more important than their

separate effects.

Several studies found no support for the deterring effects of fines. Pommerehne and

Weck-Hannemann (1996) found no impact of penalty rate in their comparison of tax

compliance in different Swiss cantons. Ali, Cecil, and Knoblett (2001) analyzed

compliance behavior of American taxpayers between 1980 and 1995. In this period penalty

rate increased from 5% to 30% of the evaded tax. Overall, increasing the fines had no

impact on compliance. However, high and low-income earners reacted differently.

Whereas low-income earners showed no change in compliance, high-income earners

reacted as expected and increased their tax payments. In experiments by Baldry (1987) and

by Webley, Robben, Elffers, and Hessing (1991) compliance was unaffected by the amount

of fines, though fines should have stronger effects in the laboratory than in the field due to

the artificial situation of gambling with the experimenters. Perceived rather than objective

severity of fines was studied in a survey by Spicer and Lundstedt (1976). Propensity for tax

evasion was not related to the perceived severity of fines.


Interestingly, increasing the fines can also have the opposite effect by initiating tax

avoidance. In a field experiment by Schwartz and Orleans (1967) a random selection of

American taxpayers received a letter emphasizing the severity of sanctions available to the

government. Reported adjusted gross income did not increase compared to previous years,

but amount of deductions increased substantially. It seems that taxpayers had tried to get

back the higher income they were forced to declare due to the threat of a fine. Fjeldstad and

Semboja (2001) report comparable results from a survey study they conducted in Tanzania.

Oppressive tax enforcement and harassment of taxpayers increased resistance to pay taxes.

Accordingly, a survey study by Strümpel (1969) revealed that unfair penalties have a

negative impact on attitudes toward the tax office and taxes.

The conclusion from the empirical evidence reported above is that deterrent effects of

fines are weak, if not negligible (as earlier reviews already had to conclude; c.f., Andreoni,

Erard, & Feinstein, 1998; Fischer, Wartick, & Mark, 1992; Pommerehne & Weck-

Hannemann, 1992). Some of the findings suggest that a policy based on deterrence is

effective only in combination with frequent audits. The most extreme penalties will have

no effect, if it is common knowledge that audits virtually do not occur.

The increasing tax avoidance and tax resistance due to an increase of fines puts into

question how fines should be assessed to be effective. On the one hand, fines should be

high enough to decrease the expected value of tax evasion and to assure its deterrent effect

on taxpayers. On the other hand, if fines are too high, the tax system would be perceived as

unjust und unfair and taxpayers would use any possibility to legally avoid their taxes. In

most countries, fines are relative to the evaded tax. In Austria, for instance, the maximum

(monetary) penalty for tax evasion is 200% of the evaded amount, though in practice the

typical sentence is a fine of 40%. However, depending on the income of the accused such a

system might yield too low fines to have deterrent effects. An alternative would be to


adjust the fine to the income of taxpayers. In an experimental survey study by

Muehlbacher, Hölzl and Kirchler (submitted manuscript) income-adjusted fines had more

impact on the sentenced taxpayer’s intention to commit the same offense again than fines

which were solely adjusted to severity of evasion.

5. Summary and discussion

The standard economic model frames the tax compliance problem as a decision under

uncertainty. Taxpayers are assumed to maximize income by weighing pros and cons of

evading taxes. Though the model provides useful tools for tax policy and its publication

stimulated a variety of tax research, empirical evidence for its validity is rather weak. Most

of its parameters have unstable and unclear effects and it is hard to draw definite

conclusions from the studies we have reviewed. If high or low-income earners are more

prone to evade their taxes remains unclear. Regarding the effects of tax rate increases, most

studies seem to support a negative relation to compliance, i.e., more tax evasion at high tax

rates. The impact of audits and fines can hardly be separated, though a combination of both

seems to be effective. Since audits are costly, however, tax policy has a wider scope in

adjusting fines.

The weak empirical support of the economic model could either be due to constraints in

the methodology applied to gather and analyze data, or the economic approach is too

narrow for a comprehensive explanation of compliance behavior.

Regarding methodology, empirical research on tax compliance faces severe problems,

mainly the problem of low external validity. Most tax experiments deal with students as

participants and often are designed as “gambling” experiments rather than providing more

realistic tax paying environments. An experiment without monetary incentives has low

external validity, but providing participants with windfall endowments additionally

enhances the mentioned gambling situation in the laboratory. Additionally, if participants’


profit is based on their behavior in the experiment, the impact of tax morale and the

possibility of other, non-financial consequences are ruled out completely. Further

methodological constraints were demonstrated in an extensive study done by Elffers,

Weigel, and Hessing (1987). The authors compared actual tax files of Dutch taxpayers with

self-reported compliance within the same sample. As expected, actual tax-evaders

frequently indicated in self-reports to be fully compliant. More surprisingly, several

participants admitted non-compliance, though their tax files proofed that income had been

honestly reported. Accordingly, experimental data and especially self-reported data have to

be interpreted very carefully before generalization.

Regarding the theoretical approach to the tax compliance problem, it might be useful to

enrich compliance research by considering alternative theories, which are based on a more

relaxed rationality assumption and take into account peoples’ actual tax behavior. Prospect

Theory (Kahneman & Tversky, 1979) for instance has been successfully applied in the

field of tax research (e.g., Elffers & Hessing, 1997; Kirchler & Maciejovsky, 2001; Yaniv,

1999). Framing effects, i.e., presenting outcomes as losses or gains, were studied

(Schepanski & Kelsey, 1990), and the withholding-phenomenon was explained by

introducing the notion of a reference point (Schepanski & Shearer, 1995). However,

Prospect Theory is of limited use for predicting behavior. For instance, it is an open

question to which reference points taxpayers naturally adopt if it is not given externally by

framing (Copeland & Cuccia, 2002).

Another research paradigm that has already been applied in tax research describes

compliance decisions as social dilemma (Dawes, 1980). According to such models,

taxpayers try to balance egoistic goals (i.e., maximizing effective income) and collective

goals (e.g., establishing health care and welfare systems). The individual is better off by

defecting and evading taxes. If too many individuals behave egoistically and evade taxes,


collective goals cannot be achieved and everyone would have been better off by

cooperating. The standard economic model fully neglects the utility of the collective.

Probably an even stronger constraint of the economic model is that tax morale is

neglected. Braithwaite (2003) suggests that taxpayers follow quit diverse motivational

postures in paying their taxes. While some may pay their taxes due to their commitment

with the community, others are disengaged or enjoy tax evasion as sort of game playing

with the state. According to Braithwaite only a small proportion of taxpayers are driven by

the latter postures and should be pursued with full rigor of the law. Therefore, Ayres and

Braithwaite (1992) argue for responsive regulation, where tax authorities treat taxpayers in

accordance to their motivational posture. Responsive regulation suggests to maintain and

stabilize compliant taxpayers and to respond with severe punishment to notorious non-

compliant taxpayers. It is likely to assume that the economic model fits best for the group

of disengaged and game playing taxpayers and loses validity among taxpayers with

different motivational postures.

A related idea was proposed in the “slippery slope” model by Kirchler, Hoelzl, and

Wahl (in press). Their approach is based on the interaction between taxpayers and

authorities, and distinguishes enforced and voluntary compliance. Whereas enforced

compliance is determined by how taxpayers perceive the power of authorities to prosecute

and punish them, voluntary compliance can be achieved if the taxpayers perceive

authorities as trustworthy and benevolent. Hence, in a climate where trust in authorities is

low, a tax policy based on audits and fines is effective. In a climate of trust, however, tax

morale, perceived fairness of the tax system, tax knowledge and social norms are important

to foster compliance. Further, a dynamic interaction of power and trust has to be

considered. Changes in trust might influence power (e.g., by whistle-blowing of compliant

citizens) and vice versa (e.g., if frequent audits are perceived as a signal of distrust). The


slippery slope model emphasizes the interaction of taxpayers and authorities, and suggests

tax authorities to move from a “cops-and-robbers” to a “service-and-client” view.

The problem of tax compliance seems much too complex to be explained by a pure

economic approach. Including alternative approaches could help to understand the

irrational behavior of taxpayers, and could expand the toolbox for an efficient tax policy.


References Ali, M. M., Cecil, H. W., & Knoblett, J. A. (2001). The effects of tax rates and

enforcement policies on taxpayer compliance: A study of self-employed

taxpayers. Atlantic Economic Journal, 29(2), 186 - 2002.

Allingham, M. G., & Sandmo, A. (1972). Income tax evasion: A theoretical analysis.

Journal of Public Economics, 1(3-4), 323-338.

Alm, J., Cronshaw, M. B., & McKee, M. (1993). Tax compliance with endogenous audit

selection-rules. Kyklos, 46(1), 27-45.

Alm, J., Jackson, B., & McKee, M. (1992). Estimating the determinants of taxpayer

compliance with experimental-data. National Tax Journal, 45(1), 107-114.

Alm, J., McClelland, G. H., & Schulze, W. D. (1992). Why do people pay taxes? Journal

of Public Economics, 48(1), 21-38.

Alm, J., Sanchez, I., & de Juan, A. (1995). Economic and noneconomic factors in tax

compliance. Kyklos, 48(1), 3-18.

Anderhub, V., Giese, S., Güth, W., Hoffmann, A., & Otto, T. (2001). Tax evasion with

earned income - an experimental study. Finanz Archiv, 58(2), 188-206.

Andreoni, J., Erard, B., & Feinstein, J. S. (1998). Tax compliance. Journal of Economic

Literature, 36(2), 818-860.

Ayres, I., & Braithwaite, J. (1992). Responsive regulation: Transcending the

deregulation debate. New York: Oxford University Press.

Baldry, J. C. (1987). Income tax evasion and the tax schedule: some experimental results.

Public Finance, 42(3), 347-383.


Boylan, S. J., & Sprinkle, G. B. (2001). Experimental evidence on the relation between

tax rates and compliance: The effect of earned vs. endowed income. The Journal

of the American Taxation Association, 23(1), 75-90.

Braithwaite, V. (2003). Dancing with tax authorities: motivational postures and non-

compliant actions. In V. Braithwaite (Ed.), Taxing democracy (pp. 1-11). Hants,

UK: Ashgate.

Christian, C. W. (1994). Voluntary compliance with the individual income tax: results

from the 1988 TCMP study. Washington, DC.

Clotfelter, C. (1983). Tax evasion and tax rates: an analysis of individual returns. The

review of economics and statistics, 65(3), 363-373.

Collins, J. H., & Plumlee, R. D. (1991). The taxpayers labor and reporting decision - the

effect of audit schemes. Accounting Review, 66(3), 559-576.

Copeland, P. V., & Cuccia, A. D. (2002). Multiple determinants of framing referents in

tax reporting and compliance. Organizational Behavior and Human Decision

Processes, 88(1), 499-526.

Dawes, R. M. (1980). Social dilemmas. Annual Review of Psychology, 31(1), 169-193.

Elffers, H., & Hessing, D. J. (1997). Influencing the prospects of tax evasion. Journal of

Economic Psychology, 18(2-3), 289-304.

Elffers, H., Weigel, R. H., & Hessing, D. J. (1987). The consequences of different

strategies for measuring tax evasion behaviour. Journal of Economic Psychology,

8(3), 311-337.

Feinstein, J. (1991). An economometric analysis of income tax evasion and its detection.

RAND Journal of Economics, 22(1), 14-35.


Fischer, C. M., Wartick, M., & Mark, M. M. (1992). Detection probability and taxpayer

compliance: a review of the literature. Journal of Accounting Literature, 11(1), 1-

46.

Fishlow, A., & Friedman, J. (1994). Tax Evasion, inflation and stabilization. Journal of

Development Economics, 43(1), 105-123.

Fjeldstad, O. H., & Semboja, J. (2001). Why people pay taxes: The case of the

development levy in Tanzania. World Development, 29(12), 2059-2074.

Friedland, N. (1982). A note on tax evasion as a function of the quality of information

about the magnitude and credibility of threatened fines: some preliminary

research. Journal of Applied Social Psychology, 12(1), 54-59.

Friedland, N., Maital, S., & Rutenberg, A. (1978). A simulation study of income tax

evasion. Journal of Public Economics, 10(1), 107-116.

Gërxhani, K., & Schram, A. (2006). Tax evasion and income source: A comparative

experimental study. Journal of Economic Psychology, 27(3), 202-222.

Guala, F., & Mittone, L. (2005). Experiments in economics: External validity and the

robustness of phenomena. Journal of Economic Methodology, 12(4), 495-515.

Kahneman, D., & Tversky, A. (1979). Prospect theory: an analysis of decision under risk.

Econometrica, 47(2), 263-291.

Kastlunger, B., Kirchler, E., Mittone, L., & Pitters, J. (submitted manuscript). Sequences

of audits, tax compliance and taxpaying strategies.

Kirchler, E. (2007). The economic psychology of tax behaviour. Cambridge: Cambridge

University Press.


Kirchler, E., Hoelzl, E., & Wahl, I. (in press). Enforced versus voluntary tax compliance:

The “slippery slope” framework. Journal of Economic Psychology.

Kirchler, E., & Maciejovsky, B. (2001). Tax compliance within the context of gain and

loss situations, expected and current asset position, and profession. Journal of

Economic Psychology, 22(2), 173-194.

Kirchler, E., Maciejovsky, B., & Schwarzenberger, H. (2005). Specious confidence after

tax audits: a contribution to the dynamics of compliance.Unpublished manuscript,

Vienna, Austria.

Kirchler, E., Muehlbacher, S., Hoelzl, E., & Webley, P. (in press). Effort and aspirations

in tax evasion: Experimental evidence. Applied Psychology: An International

Review.

Lang, O., Nöhrbaß, K.-H., & Stahl, K. (1997). On income tax avoidance: the case of

Germany. Journal of Public Economics, 66(2), 327-347.

Mason, R., & Calvin, L. D. (1978). A study of admitted income tax evasion. Law and

Society Review, 13(Fall), 73-89.

McCaffery, E. J., & Baron, J. (2004). Framing and taxation: evaluation of tax policies

involving household composition. Journal of Economic Psychology, 25(6), 679-

705.

Mittone, L. (2006). Dynamic behaviour in tax evasion: An experimental approach. The

Journal of Socio-Economics 35(5), 813-835.

Moser, D. V., Evans III, J. H., & Kim, C. K. (1995). The effects of horizontal and

exchange inequity on tax reporting decisions. Accounting Review, 70(4), 619-634.


Muehlbacher, S., Hölzl, E., & Kirchler, E. (submitted manuscript). Steuerhinterziehung

und die Berücksichtigung des Einkommmens in der Strafbemessung.

Muehlbacher, S., & Kirchler, E. (in press). Arbeitsaufwand, Anspruchsniveau und

Steuermoral. Zeitschrift für Arbeits- und Organisationspsychologie.

Park, C. G., & Hyun, J. K. (2003). Examining the determinants of tax compliance by

experimental data: a case of Korea. Journal of Policy Modeling, 25(8), 673-684.

Pommerehne, W., & Weck-Hannemann, H. (1992). Steuerhinterziehung: Einige

romantische, realistische und nicht zuletzt empirische Befunde. Zeitschrift für

Wirtschafts- und Sozialwissenschaften, 112(3), 433-466.

Pommerehne, W., & Weck-Hannemann, H. (1996). Tax rates, tax administration and

income tax evasion in Switzerland. Public Choice, 88(1-2), 161-170.

Porcano, T. M. (1988). Correlates of tax evasion. Journal of Economic Psychology, 9(1),

47-67.

Reinganum, J. F., & Wilde, L., L. (1985). Income-tax compliance in a principal agent

framework. Journal of Public Economics, 26(1), 1-18.

Roberts, M. L., Hite, P. A., & Bradley, C. F. (1994). Understanding attitudes toward

progressive taxation. Public Opinion Quarterly, 58(2), 165-190.

Schepanski, A., & Kelsey, D. (1990). Testing for framing effects in taxpayer compliance

decisions. Journal of the American Taxation Association, 12(2), 60-77.

Schepanski, A., & Shearer, T. (1995). A prospect-theory account of the income-tax

withholding phenomenon. Organizational Behavior and Human Decision

Processes, 63(2), 174-186.


Schwartz, R., & Orleans, S. (1967). On legal sanctions. University of Chicago Law

Review, 34(Winter), 274-300.

Slemrod, J. (1985). An empirical test for tax evasion. Review of Economics and Statistics,

67(2), 232-238.

Slemrod, J., Blumenthal, M., & Christian, C. (2001). Taxpayer response to an increased

probability of audit: evidence from a controlled experiment in Minnesota. Journal

of Public Economics, 79(3), 455-483.

Spicer, M. W., & Hero, R., E. (1985). Tax evasion and heuristics: a research note.

Journal of Public Economics, 26(2), 263-267.

Spicer, M. W., & Lundstedt, S. B. (1976). Understanding tax evasion. Public Finance,

21(2), 295-305.

Spicer, M. W., & Thomas, J. E. (1982). Audit probabilities and tax evasion decision: an

experimental approach. Journal of Economic Psychology, 2(3), 241-245.

Srinivasan, T. N. (1973). Tax evasion: a model. Journal of Public Economics, 2(4), 339-

346.

Strümpel, B. (1969). The contribution of survey research to public finance. In A. T.

Peacock (Ed.), Quantitative Analysis in Public Finance (pp. 13-22). New York:

Praeger.

Trivedi, V. U., Shehata, M., & Lynn, B. (2003). Impact of personal and situational factors

on taxpayer compliance: An experimental analysis. Journal of Business Ethics,

47(3), 175-197.


Trivedi, V. U., Shehata, M., & Mestelman, S. (2004). Attitudes, incentives, and tax

compliance.Unpublished manuscript, Hamilton: McMaster University,

Departement of Economics.

Tversky, A., & Kahneman, D. (1974). Judgement under uncertainty: heuristics and

biases. Science, 185, 1124-1131.

Vogel, J. (1974). Taxation and public opinion in Sweden: An interpretation of recent

survey data. National Tax Journal, 27(1), 499-513.

Wärneryd, K., & Walerud, B. (1982). Taxes and economic behaviour: some interview

data on tax evasion in Sweden. Journal of Economic Psychology, 2(3), 187-211.

Webley, P. (1987). Audit probabilities and tax evasion in a business simulation.

Economics Letters, 25(3), 267-270.

Webley, P., Robben, H. S. J., Elffers, H., & Hessing, D. J. (1991). Tax evasion: An

experimental approach. Cambridge: Cambridge University Press.

Weck-Hannemann, H., & Pommerehne, W. (1989). Einkommenssteuerhinterziehung in

der Schweiz: Eine empirische Analyse. Schweizerische Zeitschrift fur

Volkswirtschaft und Statistik, 125(4), 517-556.

Yaniv, G. (1999). Tax compliance and advance tax payments: a prospect theory analysis.

National Tax Journal, 52(4), 753-764.

Yitzhaki, S. (1974). A note on income tax evasion: a theoretical analysis. Journal of

Public Economics, 3(2), 201-202.


Documents

Why Pay Taxes? A Review of Tax Compliance Decisions Erich